(a) Field of the Invention
The present invention relates to a communication method. More particularly, the present invention relates to a communication node transmitting data via a network, and a communication method therefor.
(b) Description of the Related Art
Currently, functions for inter-networking are implemented in the Internet in a distributed fashion between a Transport Control Protocol (TCP) layer and an Internet Protocol (IP) layer. The IP layer is concerned with addressing and hop-by-hop routing for inter-networking, and the TCP layer is concerned with connection setup & tear down, flow control, error control, and congestion control for end-to-end communication.
Most main functions for communication between end hosts on the Internet are implemented in the end hosts. As for layers implemented at each end host, an IP layer for addressing and routing support for packet transfer is present on a Physical/Media Access Control (PHY/MAC) layer unique to each Access Network (AN), and a TCP layer for end-to-end error and traffic control (Instead of the TCP layer, a User Data Protocol (UDP) or Session Control Protocol (SCTP) layer may be placed) is on the IP layer, and an application layer using an interface provided by the TCP/IP is located as the upper most layer.
A router located in an access network to which an end host is connected and a backbone network connected to the access network, each include only a PHY/MAC layer and an IP layer located above the PHY/MAC layer and concerned with packet transfer. That is, a router, which is a network node, is not involved in error and traffic control with respect to packets and main control functions for communication exist only in end hosts.
In this structure, since only the end hosts have most main functions for communication, a new type of application can be introduced, just with implementation at the end hosts. As a result, the limitation of an existing telecom service, requiring the update of network functions, could be overcome, making it possible to realize the explosive growth of the Internet as now.
However, since the above structure is basically based on a simple network structure of its initial development stage, the following problems may be caused in a current or future network environment, which is different from the past environment.
First, since communication is carried out based on end hosts, the problem of scalability such as the explosive expansion of a routing table may be caused in the current or future environment where a huge number of hosts are present.
Secondly, as can be seen in the current tendency, since access networks providing access to the Internet are being continuously diversified, it may be inefficient to support such diverse access networks with a single protocol between end hosts. For example, it is expected that a variety of multiple access networks will be configured into wireless networks, rather than related art wired networks. In this case, a wired network-based TCP is insufficient to cope with such an environment.
Thirdly, an access network that is to be implemented wirelessly by various techniques, and a backbone network that is to be implemented based on an optic network with large capacity, may have different characteristics. In such a heterogeneous environment, it may be inefficient to use the structure where only end hosts are concerned with error and traffic control while a network is concerned with only packet transfer regardless of the kinds of backbone and access as in the current Internet.
Fourthly, a communication structure between end hosts in the current Internet cannot directly respond to the occurrence of problems in a network. The general principle is that a problem needs to be addressed by the nearest point. However, the communication structure between end hosts in the current Internet has limitations in that error detection and control in a network, which is the leading cause of actual error and traffic congestion, are inevitably performed in an indirect manner because only the end hosts have the functions for error and traffic control.
In order to improve the end-to-end communication structure of the current internet having the above problems, research is being conducted by studying the future Internet. However, this research is still in its conceptual stage, and an explicit method to solve the above problems has not been proposed.